Stream: git-wasmtime

Topic: wasmtime / issue #4139 Cranelift AArch64: Migrate Bitsele...

Wasmtime GitHub notifications bot (May 11 2022 at 12:30):

github-actions[bot] commented on issue #4139:

Subscribe to Label Action

cc @cfallin, @fitzgen

<details>
This issue or pull request has been labeled: "cranelift", "cranelift:area:aarch64", "cranelift:area:machinst", "cranelift:area:x64", "isle"

Thus the following users have been cc'd because of the following labels:

• cfallin: isle
• fitzgen: isle

To subscribe or unsubscribe from this label, edit the <code>.github/subscribe-to-label.json</code> configuration file.

Learn more.
</details>

Wasmtime GitHub notifications bot (May 11 2022 at 18:03):

cfallin commented on issue #4139:

cc @uweigand, looks like a bitselect failure on s390x with i8 types in one of the new runtests.

(To get this in, I would be ok with splitting narrow-type bitops into a separate runtest file and omitting target s390x until fixed)

Wasmtime GitHub notifications bot (May 11 2022 at 18:24):

akirilov-arm commented on issue #4139:

Actually before my changes the cranelift/filetests/filetests/runtests/bitops.clif file contains only one test, which IMHO is not particularly interesting (the inputs are fixed and the output does not seem to be checked), so I have been thinking of disabling the whole file for s390x (similarly to what has already been done for x86-64). Any thoughts?

Wasmtime GitHub notifications bot (May 11 2022 at 18:47):

cfallin commented on issue #4139:

That seems fine as a temporary measure. I'll file an issue to track completing support for skipped runtests on each platform in general; it's something we should definitely audit at some point.

Wasmtime GitHub notifications bot (May 12 2022 at 15:09):

uweigand commented on issue #4139:

cc @uweigand, looks like a bitselect failure on s390x with i8 types in one of the new runtests.

Oops. These were all supposed to be implemented, but it turns out the implementation of bitwise ops had a series of logic errors. Fix is here: https://github.com/bytecodealliance/wasmtime/pull/4146

Thanks for catching this!

Wasmtime GitHub notifications bot (May 13 2022 at 11:54):

akirilov-arm commented on issue #4139:

OK, I changed the test, so that it runs on all platforms except x86-64, where there is a failure.

Wasmtime GitHub notifications bot (May 16 2022 at 16:39):

cfallin commented on issue #4139:

cc @abrown for x86-64 test failure: would you mind taking a look, or at least triaging to some degree (miscompile, missing lowering, ...)?

Last updated: Nov 22 2024 at 17:03 UTC